IO8 PAUL S. WELCH. 



ing 13 different larvae, 4-6 cm. long, showed that the interval 

 spent below the water was 2 minutes and 57 seconds, and the 

 time interval spent at the top was I minute and 21 seconds. The 

 maximum period of time voluntarily spent below was 13 minutes 

 while the minimum was 20 seconds. The maximum period at 

 the top was 6 minutes and 40 seconds, and the minimum was 5 

 seconds. Comstock ('81, p. 148) states that he observed larvae 

 of this species remain under water for a half hour but in the large 

 number of observations made in this connection none remained 

 below over 13 minutes. 



In connection with the observations on the frequency of trips 

 to the surface, the question arose as to whether there was any 

 correlation between the length of the burrow and the length of 

 time between appearances at the top. In order to determine this 

 point observations were made on a number of larvae from day to 

 day as they increased the length of their burrows. Results 

 show that the length of the burrow does not in any way deter- 

 mine the time spent below the water. In fact, they show that 

 the intervals are approximately the same when the burrow is 

 but 6 cm. long as when it is over 20 cm. in length. From an 

 examination of all results secured in connection with this study, 

 the writer is convinced that the length of the periods spent under 

 water is purely a matter of supply and demand of oxygen, that 

 it is dependent upon the amount of air which the larva draws 

 into the trachea when at the top and the activity of the larva in 

 the process of feeding at the bottom of the burrow. 



Resistance to the Lack of Oxygen. Experiments were carried 

 on to determine how long larvae can live under water, when forced 

 to do so, without renewal of the air supply. The entrances to 

 the burrows were plugged in such a way that the inmates could 

 not get air at the top, and the time recorded. These petioles 

 were visited at regular intervals and the condition of the larvae 

 noted. Larvae were also placed in tubes of water which were 

 corked so that no air spaces were left at the top, and the condi- 

 tion of the specimens carefully observed. These experiments 

 showed that death usually occurred after an interval of about 

 two hours. After 15 or 20 minutes of continued submergence 

 larvae began to show the first signs of uneasiness and evidences of 

 weakness began to appear at the end of an hour. 



